-
Too Late to Train, Too Early To Use? A Study on Necessity and Viability of Low-Resource Bengali LLMs
Authors:
Tamzeed Mahfuz,
Satak Kumar Dey,
Ruwad Naswan,
Hasnaen Adil,
Khondker Salman Sayeed,
Haz Sameen Shahgir
Abstract:
Each new generation of English-oriented Large Language Models (LLMs) exhibits enhanced cross-lingual transfer capabilities and significantly outperforms older LLMs on low-resource languages. This prompts the question: Is there a need for LLMs dedicated to a particular low-resource language? We aim to explore this question for Bengali, a low-to-moderate resource Indo-Aryan language native to the Be…
▽ More
Each new generation of English-oriented Large Language Models (LLMs) exhibits enhanced cross-lingual transfer capabilities and significantly outperforms older LLMs on low-resource languages. This prompts the question: Is there a need for LLMs dedicated to a particular low-resource language? We aim to explore this question for Bengali, a low-to-moderate resource Indo-Aryan language native to the Bengal region of South Asia.
We compare the performance of open-weight and closed-source LLMs such as LLaMA-3 and GPT-4 against fine-tuned encoder-decoder models across a diverse set of Bengali downstream tasks, including translation, summarization, paraphrasing, question-answering, and natural language inference. Our findings reveal that while LLMs generally excel in reasoning tasks, their performance in tasks requiring Bengali script generation is inconsistent. Key challenges include inefficient tokenization of Bengali script by existing LLMs, leading to increased computational costs and potential performance degradation. Additionally, we highlight biases in machine-translated datasets commonly used for Bengali NLP tasks. We conclude that there is a significant need for a Bengali-oriented LLM, but the field currently lacks the high-quality pretraining and instruction-tuning datasets necessary to develop a highly effective model.
△ Less
Submitted 29 June, 2024;
originally announced July 2024.
-
Local spin dynamics in geometrically frustrated Mo pyrochlore antiferromagnet Lu$_2$Mo$_2$O$_{5-y}$N$_2$
Authors:
S. K. Dey,
K. Ishida,
H. Okabe,
M. Hiraishi,
A. Koda,
T. Honda,
J. Yamaura,
H. Kageyama,
R. Kadono
Abstract:
The magnetic ground state of oxynitride pyrochlore Lu$_2$Mo$_2$O$_{5-y}$N$_2$, a candidate compound for the quantum spin liquid ($S=1/2$, Mo$^{5+}$), was studied by muon spin rotation/relaxation experiment. In contrast to Lu$_2$Mo$_2$O$_7$ ($S=1$, Mo$^{4+}$) which exhibits the spin-glass behavior with a freezing temperature $T_g\simeq16$ K, no such spin freezing or long range magnetic order was ob…
▽ More
The magnetic ground state of oxynitride pyrochlore Lu$_2$Mo$_2$O$_{5-y}$N$_2$, a candidate compound for the quantum spin liquid ($S=1/2$, Mo$^{5+}$), was studied by muon spin rotation/relaxation experiment. In contrast to Lu$_2$Mo$_2$O$_7$ ($S=1$, Mo$^{4+}$) which exhibits the spin-glass behavior with a freezing temperature $T_g\simeq16$ K, no such spin freezing or long range magnetic order was observed down to 0.3 K. Moreover, two separate magnetic domains were detected below $\sim$13 K, which were characterized by differences in spin dynamics. The first is the "sporadic" spin fluctuation seen in frustrated antiferromagnets, where the amplitude of the hyperfine fields suggests that the excitation comprises a local cluster of unpaired spins. The other is rapid paramagnetic fluctuation, which is only weakly suppressed at low temperatures. In place of the paramagnetic fluctuation, the volume fraction for the sporadic fluctuation steadily increases with decreasing temperature, indicating the formation of an excitation gap with a broad distribution of the gap energy involving null gap.
△ Less
Submitted 7 September, 2022; v1 submitted 27 June, 2022;
originally announced June 2022.
-
Multimodal Hate Speech Detection from Bengali Memes and Texts
Authors:
Md. Rezaul Karim,
Sumon Kanti Dey,
Tanhim Islam,
Md. Shajalal,
Bharathi Raja Chakravarthi
Abstract:
Numerous machine learning (ML) and deep learning (DL)-based approaches have been proposed to utilize textual data from social media for anti-social behavior analysis like cyberbullying, fake news detection, and identification of hate speech mainly for highly-resourced languages such as English. However, despite having a lot of diversity and millions of native speakers, some languages like Bengali…
▽ More
Numerous machine learning (ML) and deep learning (DL)-based approaches have been proposed to utilize textual data from social media for anti-social behavior analysis like cyberbullying, fake news detection, and identification of hate speech mainly for highly-resourced languages such as English. However, despite having a lot of diversity and millions of native speakers, some languages like Bengali are under-resourced, which is due to a lack of computational resources for natural language processing (NLP). Similar to other languages, Bengali social media contents also include images along with texts (e.g., multimodal memes are posted by embedding short texts into images on Facebook). Therefore, only the textual data is not enough to judge them since images might give extra context to make a proper judgement. This paper is about hate speech detection from multimodal Bengali memes and texts. We prepared the only multimodal hate speech dataset for-a-kind of problem for Bengali, which we use to train state-of-the-art neural architectures (e.g., Bi-LSTM/Conv-LSTM with word embeddings, ConvNets + pre-trained language models, e.g., monolingual Bangla BERT, multilingual BERT-cased/uncased, and XLM-RoBERTa) to jointly analyze textual and visual information for hate speech detection. Conv-LSTM and XLM-RoBERTa models performed best for texts, yielding F1 scores of 0.78 and 0.82, respectively. As of memes, ResNet-152 and DenseNet-161 models yield F1 scores of 0.78 and 0.79, respectively. As for multimodal fusion, XLM-RoBERTa + DenseNet-161 performed the best, yielding an F1 score of 0.83. Our study suggests that text modality is most useful for hate speech detection, while memes are moderately useful.
△ Less
Submitted 21 December, 2022; v1 submitted 19 April, 2022;
originally announced April 2022.
-
Use of Digital Technologies in Public Health Responses to Tackle Covid-19: the Bangladesh Perspective
Authors:
Samrat Kumar Dey,
Khaleda Mehrin,
Lubana Akter,
Mshura Akter
Abstract:
This paper aims to study the fight against COVID-19 in Bangladesh and digital intervention initiatives. To achieve the purpose of our research, we conducted a methodical review of online content. We have reviewed the first digital intervention that COVID-19 has been used to fight against worldwide. Then we reviewed the initiatives that have been taken in Bangladesh. Our paper has shown that while…
▽ More
This paper aims to study the fight against COVID-19 in Bangladesh and digital intervention initiatives. To achieve the purpose of our research, we conducted a methodical review of online content. We have reviewed the first digital intervention that COVID-19 has been used to fight against worldwide. Then we reviewed the initiatives that have been taken in Bangladesh. Our paper has shown that while Bangladesh can take advantage of the digital intervention approach, it will require rigorous collaboration between government organizations and universities to get the most out of it. Public health can become increasingly digital in the future, and we are reviewing international alignment requirements. This exploration also focused on the strategies for controlling, evaluating, and using digital technology to strengthen epidemic management and future preparations for COVID-19.
△ Less
Submitted 10 March, 2022;
originally announced March 2022.
-
MirrorME: Implementation of an IoT based Smart Mirror through Facial Recognition and Personalized Information Recommendation Algorithm
Authors:
Khandaker Mohammad Mohi Uddin,
Samrat Kumar Dey,
Gias Uddin Parvez,
Ayesha Siddika Mukta,
Uzzal Kumar Acharjee
Abstract:
We are living in the era of the fourth industrial revolution, which also treated as 4IR or Industry 4.0. Generally, 4IR considered as the mixture of robotics, artificial intelligence (AI), quantum computing, the Internet of Things (IoT) and other frontier technologies. It is obvious that nowadays a plethora of smart devices is providing services to make the daily life of human easier. However, in…
▽ More
We are living in the era of the fourth industrial revolution, which also treated as 4IR or Industry 4.0. Generally, 4IR considered as the mixture of robotics, artificial intelligence (AI), quantum computing, the Internet of Things (IoT) and other frontier technologies. It is obvious that nowadays a plethora of smart devices is providing services to make the daily life of human easier. However, in the morning most people around the globe use a traditional mirror while preparing themselves for daily task. The aim is to build a low-cost intelligent mirror system that can display a variety of details based on user recommendations. Therefore, in this article, Internet of Things (IoT) and AI-based smart mirror is introduced that will support the users to receive the necessary daily update of weather information, date, time, calendar, to-do list, updated news headlines, traffic updates, COVID-19 cases status and so on. Moreover, a face detection method also implemented with the smart mirror to construct the architecture more secure. Our proposed MirrorME application provides a success rate of nearly 87% in interacting with the features of face recognition and voice input. The mirror is capable of delivering multimedia facilities while maintaining high levels of security within the device.
△ Less
Submitted 28 February, 2021;
originally announced March 2021.
-
Affective State Recognition through EEG Signals Feature Level Fusion and Ensemble Classifier
Authors:
Md. Mahbubur Rahman,
Akash Poddar,
Md. Golam Rabiul Alam,
Samrat Kumar Dey
Abstract:
Human affects are complex paradox and an active research domain in affective computing. Affects are traditionally determined through a self-report based psychometric questionnaire or through facial expression recognition. However, few state-of-the-arts pieces of research have shown the possibilities of recognizing human affects from psychophysiological and neurological signals. In this article, el…
▽ More
Human affects are complex paradox and an active research domain in affective computing. Affects are traditionally determined through a self-report based psychometric questionnaire or through facial expression recognition. However, few state-of-the-arts pieces of research have shown the possibilities of recognizing human affects from psychophysiological and neurological signals. In this article, electroencephalogram (EEG) signals are used to recognize human affects. The electroencephalogram (EEG) of 100 participants are collected where they are given to watch one-minute video stimuli to induce different affective states. The videos with emotional tags have a variety range of affects including happy, sad, disgust, and peaceful. The experimental stimuli are collected and analyzed intensively. The interrelationship between the EEG signal frequencies and the ratings given by the participants are taken into consideration for classifying affective states. Advanced feature extraction techniques are applied along with the statistical features to prepare a fused feature vector of affective state recognition. Factor analysis methods are also applied to select discriminative features. Finally, several popular supervised machine learning classifier is applied to recognize different affective states from the discriminative feature vector. Based on the experiment, the designed random forest classifier produces 89.06% accuracy in classifying four basic affective states.
△ Less
Submitted 14 February, 2021;
originally announced February 2021.
-
DeepHateExplainer: Explainable Hate Speech Detection in Under-resourced Bengali Language
Authors:
Md. Rezaul Karim,
Sumon Kanti Dey,
Tanhim Islam,
Sagor Sarker,
Mehadi Hasan Menon,
Kabir Hossain,
Bharathi Raja Chakravarthi,
Md. Azam Hossain,
Stefan Decker
Abstract:
The exponential growths of social media and micro-blogging sites not only provide platforms for empowering freedom of expressions and individual voices, but also enables people to express anti-social behaviour like online harassment, cyberbullying, and hate speech. Numerous works have been proposed to utilize textual data for social and anti-social behaviour analysis, by predicting the contexts mo…
▽ More
The exponential growths of social media and micro-blogging sites not only provide platforms for empowering freedom of expressions and individual voices, but also enables people to express anti-social behaviour like online harassment, cyberbullying, and hate speech. Numerous works have been proposed to utilize textual data for social and anti-social behaviour analysis, by predicting the contexts mostly for highly-resourced languages like English. However, some languages are under-resourced, e.g., South Asian languages like Bengali, that lack computational resources for accurate natural language processing (NLP). In this paper, we propose an explainable approach for hate speech detection from the under-resourced Bengali language, which we called DeepHateExplainer. Bengali texts are first comprehensively preprocessed, before classifying them into political, personal, geopolitical, and religious hates using a neural ensemble method of transformer-based neural architectures (i.e., monolingual Bangla BERT-base, multilingual BERT-cased/uncased, and XLM-RoBERTa). Important(most and least) terms are then identified using sensitivity analysis and layer-wise relevance propagation(LRP), before providing human-interpretable explanations. Finally, we compute comprehensiveness and sufficiency scores to measure the quality of explanations w.r.t faithfulness. Evaluations against machine learning~(linear and tree-based models) and neural networks (i.e., CNN, Bi-LSTM, and Conv-LSTM with word embeddings) baselines yield F1-scores of 78%, 91%, 89%, and 84%, for political, personal, geopolitical, and religious hates, respectively, outperforming both ML and DNN baselines.
△ Less
Submitted 6 August, 2021; v1 submitted 28 December, 2020;
originally announced December 2020.
-
High spin states of $^{204}$At: isomeric states and shears band structure
Authors:
D. Kanjilal,
S. K. Dey,
S. S. Bhattacharjee,
A. Bisoi,
M. Das,
C. C. Dey,
S. Nag,
R. Palit,
S. Ray,
S. Saha,
J. Sethi,
S. Saha
Abstract:
High-spin states of neutron deficient Trans-Lead nucleus $^{204}$At were populated up to $\sim 8\,{\rm MeV}$ excitation through the $^{12}$C + $^{197}$Au fusion evaporation reaction. Decay of the associated levels through prompt and delayed $γ$-ray emissions were studied to evaluate the underlying nuclear structure. The level scheme, which was partly known, was extended further. An isomeric…
▽ More
High-spin states of neutron deficient Trans-Lead nucleus $^{204}$At were populated up to $\sim 8\,{\rm MeV}$ excitation through the $^{12}$C + $^{197}$Au fusion evaporation reaction. Decay of the associated levels through prompt and delayed $γ$-ray emissions were studied to evaluate the underlying nuclear structure. The level scheme, which was partly known, was extended further. An isomeric $16^+$ level with observed lifetime $τ=52 \pm 5\, {\rm ns}$, was established from our measurements. Attempts were made to interpret the excited states based on multi quasiparticle and hole structures involving $2f_{5/2}$, $1h_{9/2}$, and $1i_{13/2}$ shell model states, along with moderate core excitation. Magnetic dipole band structure over the spin parity range:~$16^+ - 23^+$ was confirmed and evaluated in more detail, including the missing cross-over $E2$ transitions. Band-crossing along the shears band was observed and compared with the evidence of similar phenomena in the neighbouring neutron deficient $^{202}$Bi, $^{205}$Rn isotones and the $^{203}$At isotope. Based on comparison of the measured $B(M1)/B(E2)$ values for transitions along the band with the semiclassical model based estimates, the shears band of $^{204}$At was established along with the level scheme.
△ Less
Submitted 1 September, 2022; v1 submitted 12 June, 2020;
originally announced June 2020.
-
Advanced Single Image Resolution Upsurging Using a Generative Adversarial Network
Authors:
Md. Moshiur Rahman,
Samrat Kumar Dey,
Kabid Hassan Shibly
Abstract:
The resolution of an image is a very important criterion for evaluating the quality of the image. A higher resolution of an image is always preferable as images of lower resolution are unsuitable due to fuzzy quality. A higher resolution of an image is important for various fields such as medical imaging; astronomy works and so on as images of lower resolution becomes unclear and indistinct when t…
▽ More
The resolution of an image is a very important criterion for evaluating the quality of the image. A higher resolution of an image is always preferable as images of lower resolution are unsuitable due to fuzzy quality. A higher resolution of an image is important for various fields such as medical imaging; astronomy works and so on as images of lower resolution becomes unclear and indistinct when their sizes are enlarged. In recent times, various research works are performed to generate a higher resolution of an image from its lower resolution. In this paper, we have proposed a technique of generating higher resolution images form lower resolution using Residual in Residual Dense Block network architecture with a deep network. We have also compared our method with other methods to prove that our method provides better visual quality images.
△ Less
Submitted 30 May, 2020;
originally announced June 2020.
-
Phase components in Zr7Ni10 and Hf7Ni10 binary alloys; investigations by perturbed angular correlation spectroscopy and first principles calculations
Authors:
S. K. Dey,
C. C. Dey,
S. Saha,
G. Bhattacharjee,
D. Banerjee,
D. Toprek
Abstract:
Intermetallic compounds Zr7Ni10 and Hf7Ni10 have been studied by perturbed angular correlation (PAC) spectroscopy considering the fact that Zr7Ni10 has application as hydrogen storage material in fuel cell. In stoichiometric Zr7Ni10, the phase Zr2Ni7 is found to be dominant (~38%) while a fraction of ~25% is found for the Zr7Ni10 phase at room temperature. In this compound, a phase due to Zr8Ni21…
▽ More
Intermetallic compounds Zr7Ni10 and Hf7Ni10 have been studied by perturbed angular correlation (PAC) spectroscopy considering the fact that Zr7Ni10 has application as hydrogen storage material in fuel cell. In stoichiometric Zr7Ni10, the phase Zr2Ni7 is found to be dominant (~38%) while a fraction of ~25% is found for the Zr7Ni10 phase at room temperature. In this compound, a phase due to Zr8Ni21 (~10%) is found from room temperature up to 773 K but, this is not found at 873 K and above. In the stoichiometric Hf7Ni10 sample, the phase due to Hf7Ni10 is found as a minor phase (~22%) at room temperature. In this system, no phase of Hf2Ni7 is observed but, a different phase due to HfNi3 is found to be dominant (~62%). It is found that the site fraction of Hf7Ni10 enhances with temperature at the expense of HfNi3 and this phase becomes predominant (~57%) at 673 K and above. The change of phase fractions of HfNi3 and Hf7Ni10 with temperature is found to be reversible. The phase components in Zr7Ni10 and Hf7Ni10 have been determined also from X-ray powder diffraction (XRD) and transmission electron microscopy/selected area electron diffraction (TEM/SAED) measurements. Ab-initio calculations using the all electron full potential (linearized) augmented plane wave [FP-(L)APW] method, within the framework of the density functional theory (DFT), have been performed to determine the electric field gradients at the 181Ta impurity sites and therefrom to assign the different components observed from PAC measurements. Present experimental and calculated results of EFG support the Cmca space group for both Zr7Ni10 and Hf7Ni10 compounds.
△ Less
Submitted 20 September, 2018;
originally announced September 2018.
-
Crystalline phases in Zr9Ni11 and Hf9Ni11 intermetallics; Investigations by perturbed angular correlation spectroscopy and ab initio calculations
Authors:
S. K. Dey,
C. C. Dey,
S. Saha,
G. Bhattacharjee,
J. Belošević-Čavor,
D. Toprek
Abstract:
Crystalline phases formed in stoichiometric Zr$_9$Ni$_{11}$ and Hf$_9$Ni$_{11}$ have been studied by perturbed angular correlation (PAC) spectroscopy, XRD and TEM/SAED measurements. In Zr$_9$Ni$_{11}$, the phases Zr$_9$Ni$_{11}$ ($\sim$89\%) and Zr$_8$Ni$_{21}$ ($\sim$11\%) have been found at room temperature from PAC measurements. At 773 K, Zr$_9$Ni$_{11}$ partially decomposes to Zr$_7$Ni$_{10}$…
▽ More
Crystalline phases formed in stoichiometric Zr$_9$Ni$_{11}$ and Hf$_9$Ni$_{11}$ have been studied by perturbed angular correlation (PAC) spectroscopy, XRD and TEM/SAED measurements. In Zr$_9$Ni$_{11}$, the phases Zr$_9$Ni$_{11}$ ($\sim$89\%) and Zr$_8$Ni$_{21}$ ($\sim$11\%) have been found at room temperature from PAC measurements. At 773 K, Zr$_9$Ni$_{11}$ partially decomposes to Zr$_7$Ni$_{10}$ and at 973 K, it is completely decomposed to ZrNi and Zr$_7$Ni$_{10}$. In Hf$_9$Ni$_{11}$, a predominant phase ($\sim$81\%) due to HfNi is found at room temperature while the phase Hf$_9$Ni$_{11}$ is produced as a minor phase ($\sim$19\%). No compositional phase change at higher temperature is found in Hf$_9$Ni$_{11}$. Phase components found from XRD and TEM/SAED measurements are similar to those observed from PAC measurements. Electric field gradients in Zr$_9$Ni$_{11}$ and Hf$_9$Ni$_{11}$ have been calculated by density functional theory (DFT) using all electron full potential (linearized) augmented plane wave plus local orbitals [FP-(L)APW+lo] method in order to assign the phase components.
△ Less
Submitted 26 October, 2018; v1 submitted 12 December, 2017;
originally announced December 2017.
-
Identification of phase components in Zr-Ni and Hf-Ni intermetallic compounds; Investigations by perturbed angular correlation spectroscopy and first principles calculations
Authors:
S. K. Dey,
C. C. Dey,
S. Saha,
J. Belosevic-Cavor,
D. Toprek
Abstract:
Time-differential perturbed angular correlation (TDPAC) measurements have been carried out in stoichiometric ZrNi$_3$ and HfNi$_3$ intermetallic compounds using $^{181}$Ta probe in the temperature range 77-1073 K considering the immense technological applications of Zr-Ni and Hf-Ni intermetallic compounds. In ZrNi$_3$, four components due to the production of Zr$_2$Ni$_7$, Zr$_8$Ni$_{21}$, Zr$_7$N…
▽ More
Time-differential perturbed angular correlation (TDPAC) measurements have been carried out in stoichiometric ZrNi$_3$ and HfNi$_3$ intermetallic compounds using $^{181}$Ta probe in the temperature range 77-1073 K considering the immense technological applications of Zr-Ni and Hf-Ni intermetallic compounds. In ZrNi$_3$, four components due to the production of Zr$_2$Ni$_7$, Zr$_8$Ni$_{21}$, Zr$_7$Ni$_{10}$ and ZrNi$_3$ have been found at room temperature. The HfNi$_3$ sample produces five electric quadrupole interaction frequencies at room temperature. The phase HfNi$_3$ is strongly produced in stoichiometric sample of HfNi$_3$ where two non-equivalent Hf sites are found to be present. Besides this phase, two other phases due to Hf$_2$Ni$_7$ and Hf$_8$Ni$_{21}$ have been found but, we do not observe any phase due to Hf$_7$Ni$_{10}$. X-ray diffraction, TEM/energy dispersive X-ray spectroscopy (EDX) and TEM-selected area electron diffraction (SAED) measurements were used to further characterize the investigated materials and it was found that these results agree with the TDPAC results. In order to confirm findings from TDPAC measurements, density functional theory (DFT) based calculations of electric field gradients (EFG) and asymmetry parameters at the sites of $^{181}$Ta probe nucleus were performed. Our calculated results are found to be in excellent agreement with the experimental results.
△ Less
Submitted 20 September, 2018; v1 submitted 29 March, 2017;
originally announced March 2017.
-
Electric field gradients in Zr$_8$Ni$_{21}$ and Hf$_8$Ni$_{21}$ intermetallic compounds; Results from perturbed angular correlation measurements and first-principles density functional theory
Authors:
S. K. Dey,
C. C. Dey,
S. Saha,
J. Belošević-Čavor
Abstract:
Numerous technological applications of Ni-based Zr and Hf intermetallic alloys promoted comprehensive studies in Zr$_8$Ni$_{21}$ and Hf$_8$Ni$_{21}$ by perturbed angular correlation (PAC) spectroscopy, which were not studied earlier until this report. The different phases produced in the samples have been identified by PAC and X-ray diffraction (XRD) measurements. Using $^{181}$Hf probe, two non-e…
▽ More
Numerous technological applications of Ni-based Zr and Hf intermetallic alloys promoted comprehensive studies in Zr$_8$Ni$_{21}$ and Hf$_8$Ni$_{21}$ by perturbed angular correlation (PAC) spectroscopy, which were not studied earlier until this report. The different phases produced in the samples have been identified by PAC and X-ray diffraction (XRD) measurements. Using $^{181}$Hf probe, two non-equivalent Zr/Hf sites have been observed in both Zr$_8$Ni$_{21}$ and Hf$_8$Ni$_{21}$ compounds. From present PAC measurements in Zr$_8$Ni$_{21}$, a component due to the production of Zr$_7$Ni$_{10}$ by eutectic reaction from the liquid metals is also observed. The phase Zr$_7$Ni$_{10}$, however, is not found from the XRD measurement. In Zr$_8$Ni$_{21}$, while the results do not change appreciably up to 973 K exhibit drastic changes at 1073 K. In Hf$_8$Ni$_{21}$, similar results for the two non-equivalent sites have been found but site fractions are in reverse order. In this alloy, a different contaminating phase, possibly due to HfNi$_3$, has been found from PAC measurements but is not found from XRD measurement. Density functional theory (DFT) based calculations of electric field gradient (EFG) and asymmetry parameter ($η$) at the sites of $^{181}$Ta probe nucleus allowed us to assign the observed EFG fractions to the various lattice sites in (Zr/Hf)$_8$Ni$_{21}$ compounds.
△ Less
Submitted 29 September, 2018; v1 submitted 5 October, 2016;
originally announced October 2016.
-
Algorithm For Multi-Hand Finger Counting: An Easy Approach
Authors:
Sumit Kumar Dey,
Shubham Anand
Abstract:
In this paper we propose an easy algorithm for real time hand finger counting involving one or more than one hand. Hand finger counting is a simple medium for Human-Computer Interface which can prove to be a convenient input method for driving interactive menus, small applications and interactive games. Here, we also calculate the orientation of the hand which can be used to provide inputs for the…
▽ More
In this paper we propose an easy algorithm for real time hand finger counting involving one or more than one hand. Hand finger counting is a simple medium for Human-Computer Interface which can prove to be a convenient input method for driving interactive menus, small applications and interactive games. Here, we also calculate the orientation of the hand which can be used to provide inputs for the direction and/or speed control of a robot, controlling mouse cursor or slide-show presentations. If being used by a single person, with his or her two hands, the person can trigger ten different commands with fingers in addition to the orientation of the hand. We tried to use very simple algorithm using very basic mathematics of 2D coordinate geometry and avoided the conventional use of calculus, contours and convex hull. Anyone seeking for an easy to implement hand finger counting algorithm can refer to it.
△ Less
Submitted 10 April, 2014;
originally announced April 2014.
-
Optical Properties and Raman Studies of Partially Edge Terminated Vertically Aligned Nanocrystalline MoS2 Thin Film
Authors:
Anand P. S. Gaur,
Satyaprakash Sahoo,
Majid Ahmadi,
Maxime J-F Guinel,
Sanjeev K. Gupta,
Ravindra Pandey,
Sandwip K Dey,
Ram S. Katiyar
Abstract:
The optical and vibrational properties of nanocrystalline thin films of MoS2, comprised of a mixture of edge terminated vertically aligned (ETVA) and (001)-oriented regions, on large insulating substrates are reported. From high resolution transmission electron microscopy (HRTEM), the average size of ETVA nanocrystals were ~5 nm and each nanocrystal consisted of only 3 to 5 monolayers of MoS2. The…
▽ More
The optical and vibrational properties of nanocrystalline thin films of MoS2, comprised of a mixture of edge terminated vertically aligned (ETVA) and (001)-oriented regions, on large insulating substrates are reported. From high resolution transmission electron microscopy (HRTEM), the average size of ETVA nanocrystals were ~5 nm and each nanocrystal consisted of only 3 to 5 monolayers of MoS2. The films were highly transparent (~80%) but the percent of transmittance decreased as the energy of the incident light approached to the band gap. Additionally, weak excitonic peaks were observed both in the absorption and transmission spectra. The room temperature Raman study showed that both the E12g and A1g modes were significantly broader, and a few additional Raman modes were observed when compared to bulk MoS2. The broadening of the A1g mode was analyzed using the phonon-confinement model and the calculated particle size was in good agreement with TEM observations. Moreover, the temperature coefficient of the A1g mode was estimated from the temperature dependent Raman studies.
△ Less
Submitted 18 July, 2013;
originally announced July 2013.
-
Self-regulated charge transfer and band tilt in nm-scale polar GaN films
Authors:
M. -H Tsai,
S. K. Dey
Abstract:
To date, the generic polarization of Bernardini, Fiorentini and Vanderbilt (PBFV) has been widely used to address the issue of polarity in III-V nitride semiconductors, but improvements in nitride materials and the performance of optoelectronic devices have been limited. The current first-principles calculation for the electronic structures of nm-scale [0001] GaN films show that the internal ele…
▽ More
To date, the generic polarization of Bernardini, Fiorentini and Vanderbilt (PBFV) has been widely used to address the issue of polarity in III-V nitride semiconductors, but improvements in nitride materials and the performance of optoelectronic devices have been limited. The current first-principles calculation for the electronic structures of nm-scale [0001] GaN films show that the internal electric fields and the band tilt of these films are in opposite direction to those predicted by PBFV. Additionally, it is determined that an intrinsic self-regulated charge transfer across the film limits the electrostatic potential difference across the film, which renders the local conduction band energy minimum (at the Ga-terminated surface) approximately equal to the local valence band energy maximum (at the N-terminated surface). This effect is found to occur in films thicker than ~4nm.
△ Less
Submitted 13 April, 2006;
originally announced April 2006.
